Evolutionary Dynamics in Stock Market: An Empirical Investigation for Chinese-A share Markets

This thesis investigates the applicability of evolutionary finance theories and behavioural biases in explaining stock price dynamics in the Chinese A-share markets. We extend the Efficient Market Hypothesis (EMH), which states that if markets are efficient, future prices cannot be forecast, by incorporating evolutionary finance models and behavioural finance insights. Specifically, we examine the validity of evolutionary dynamic models and detect behavioural biases, addressing critical gaps in understanding investor behaviours and market dynamics. The unique nature of the Chinese A-share market, predominantly open to domestic investors, provides a compelling context to explore how evolutionary dynamics and behavioural biases manifest and influence market outcomes. The empirical analysis begins with the application of Kaldasch’s (2014) evolutionary economic model and Hommes’ (2000) adaptive evolutionary system framework to so as to detect patterns of stock price evolution over both the short-term and the long-term periods in the Chinese A-share markets. Results demonstrate that stock returns generally follow a Laplace-Gaussian mixture distribution, validating the evolutionary finance approach and highlighting the significance of heterogeneous investor interactions. Recognising that relying solely on EMH and evolutionary dynamics could lead to biased results, this thesis further empirically assesses the presence and impact of prevalent behavioural biases: managerial overconfidence, and herding. Adopting the crash-risk framework developed by Kim et al. (2016) and extended by Li et al. (2025), we find clear evidence of managerial overconfidence in the Chinese stock market. The results show that industry-level managerial overconfidence is positively associated with future stock price crash risk, indicating that optimistic managerial beliefs may contribute to greater downside tail risk through aggressive corporate decisions and the delayed disclosure of adverse information. The findings also suggest that this relationship becomes more significant in the post-COVID period, when uncertainty and information inefficiency appear to have strengthened. Further cross-sectional analysis reveals that effect of overconfidence is not evenly distributed within industries, but is concentrated primarily in Manufacturing, where managerial overconfidence shows a significant positive relationship with future crash risk under both crash risk measures. Additional tests on the mediating factor of inefficient investment suggest that, although inefficient investment does not appear to be a strong mediation at the aggregated industry level, it remains relevant in specific sectors, particularly Manufacturing Industry. The herding behaviour analysis utilises the generalised Cross-sectional Standard Deviation (CSSD) method developed by Yao et al. (2014), enhancing its detection power by incorporating turnover and dummy variables for the market volume. This methodology confirms the existence of herding among investors, especially pronounced at the industry level and during volatile market conditions, although the evidence suggests diminishing effects over time. Overall, this research consistently and significantly contributes to the evolutionary finance literature by offering delicate insights into investor behaviour in emerging markets. In practice, the findings provide valuable guidance for policymakers, financial practitioners, and market participants, aiding in the development of improved market efficiency, regulatory frameworks, and more informed investment decision-making strategies in the Chinese A-share markets. Policymakers should recognise the persistence of behavioural biases, indicating the need of integrating behavioural perspectives into financial regulation. Financial practitioners, including analysts and institutional investors, can use the findings to develop more effective investment strategies and risk management practices that account for prevalent behavioural biases. Market participants, particularly individual investors, benefit from increased awareness of their biases, enabling better informed decisions and potentially enhancing personal investment outcomes.

Name dc:type.qualificationname

Ph.D.

Level dc:type.qualificationlevel

PhD thesis

Grantor dc:publisher.institution

University of Westminster

Year dc:date.issued

2025

Author dc:creator

• Huang, Zichun

Advisors dc:contributor.advisor

• van Dellen, S.
• Li, X.
• Thapar, H.

Identifier

oai:westminsterresearch.westminster.ac.uk:x64yv

OAI identifier oai:identifier

oai:westminsterresearch.westminster.ac.uk:x64yv